In color image processing applications, it is useful to understand how humans perceive colors. By understanding the human visual system and its sensitivity to certain colors, one can more effectively create and manipulate images to create a desired visual effect. This assertion is particularly true in image processing applications that intentionally alter an image to perform a desired function, like hiding information in an image or compressing an image. In digital watermarking, for example, one objective is to encode auxiliary information into a signal, such as an image or video sequence, so that the auxiliary information is substantially imperceptible to humans in an output form of the signal. Similarly, image compression applications seek to decrease the amount of data required to represent an image without introducing noticeable artifacts.
One aspect of present disclosure relates to selective color masking of images. One example is a method for mapping a change in an image attribute such as luminance or chrominance to a change in color components such that the change is less visible. This form of color masking is particularly useful in watermark encoding applications, but applies to other applications as well.
A color masking method may be incorporated into a watermark encoder to reduce visibility of a watermark. The watermark encodes auxiliary information in an image by modifying attributes of image samples in the image. As part of the encoding process, the encoder computes a change in an image attribute, such as luminance, to an image sample to encode auxiliary information in the image. It changes color values of the image sample to effect the change in luminance with minimized impact on visibility.
In one implementation, a mapping process transforms a change in an image sample attribute, such as luminance, to a change in color components of the image sample based on the color values of the image sample. The mapping process may be implemented with a look up table, where the image sample color values are used to look up a corresponding change in color values. For images represented as an array of color vectors (e.g., color triplets like Red Green Blue or Cyan Magenta Yellow), the look up table may be implemented as a multidimensional look up table with color vectors used to index a corresponding change in the color values of the image samples. The mapping process may transform a change in an image sample attribute to an absolute change in the color values of the sample, or to a relative change, such as a scale factor. To effect a change in an image sample, for example, the mapping process uses the color of an image sample to look up a scale factor for each of the color values of the image sample. It uses this scale factor, along with the desired change in the image attribute, to compute an absolute change in the color values.
Another aspect of the disclosure is a color masking method that may be used in digital watermark and other applications. The method reads the color values of an image sample and a corresponding change of an attribute of the image sample. Based on the color values of the image sample, the method maps the change in the image sample attribute to a change in color components of the image sample. The change in the color components is equivalent to the change in the image sample attribute, yet reduces visibility of the change for the specific color values of the image sample. The image sample attribute may be luminance, chrominance, or some other attribute.
Another aspect of the disclosure is the use of a color key to determine validity of a watermark in an image. The color key indicates how an attribute of image sample is changed to encode a watermark based on the color of that image sample. Since a watermark encoded using this color key is dependent on the color key and the specific color values of the image in which it is embedded, the color key can be used to check whether a watermark is valid in an image suspected of containing a watermark. For example, if the watermark is forged without the knowledge of the key or copied from a different image, then the decoder is likely to determine that the watermark has not been encoded in the color channels specified by the color key. An invalid watermark will likely be weak in the color channels specified by the key. As such, invalid watermarks found in the wrong color channels or having a weak signal strength in the correct color channels as indicated by the key are considered to be invalid.
Another aspect of the disclosure is a user interface method that enables a user to control the strength of a watermark in specified colors or color regions of an image. The user interface enables the user to specify the color or color regions to embed a watermark signal more or less strongly than other colors. The transition into selected color regions is made less visible, by smoothly changing the signal strength depending on the distance from the selected color region. Also, it enables the user to select the color region by selecting pixels having the desired color in the image to be watermarked.
The color masking technology outlined above applies to image processing operations performed in the spatial domain as well as in other transform domains. For example, another aspect of the disclosure is a method that performs color adaptive encoding of auxiliary data in transform coefficients of the image. In this method, image blocks are transformed into transform coefficients, which are then altered to encode auxiliary data. The alterations are adaptive to the color of the image because they are dependent on the characteristic color of the block to which they are made. In one implementation, for example, the average color of the block is used to look up the corresponding color channels in which to embed the watermark.
Another aspect of the disclosure is method including: determining a plurality of color attributes associated with video or imagery; determining which samples representing the video or imagery should receive digital watermarking based on the plurality of color attributes; transforming at least some of the samples into a transform domain; and utilizing a programmed electronic processor, modifying transform domain coefficients representing the samples to hide the digital watermarking therein.
Still another aspect is an apparatus comprising: electronic memory for storing a plurality of color attributes associated with video or imagery; and an electronic processor. The electronic processor is programmed for: determining which samples representing the video or imagery should receive digital watermarking based on the plurality of color attributes; transforming at least some of the samples into a transform domain; and utilizing a programmed electronic processor, modifying transform domain coefficients representing the samples to hide the digital watermarking therein.
Further features of the disclosure will become apparent with reference to the following detailed description and accompanying drawings.